Checking for existing records
When data is added to iBase as a part of an import, it might match existing records. As a part of the import specification, you can choose how potential matches are handled.
About this task
Depending on whether you are importing entities or links, the options that you are provided have the same basic choices, but if you are importing links, you must make decisions not only for the link record, but also for the entities that are at either end of that link.
Procedure
For each type of record that you are creating, to check for matching records:
Select the option to check for matching records of this type using identifiers, and in the case of links, link ends.
Choose the method of handling potential matches:
Option
Description
Update it
Update the matching record with the supplied information.
Don't update it
Discard data that matches existing records in the database.
Confirm action
Each time matching data is identified, choose whether to update the record.
Optional: You can use the Treat identifiers without values as matches option to determine whether or not fields marked as identifiers should count as matches, if they do not contain values. The option is checked by default which means that identifiers that do not contain values are treated as a match.
It is important to understand how this option can change matching behavior. In this example you are working with Person records and have First Name, Last Name and Date of Birth marked as identifers. You set matches to 'update if found', 'create if not found', and leave the identifiers without values option checked. When you run the import, records with matching identifiers are updated, even if the Date of Birth field, for example, is blank. If you then uncheck the identifiers without values option and run the import again, a new record is created even though the first and last names are a match. The importer regards the blank value in Date of Birth as not matching, and does not update the record.